python - Design and structure of batch/group operation functions -

assuming want design function in python takes input one-dimensional (1d) list, processes list , returns processed result.

 def process_list(li):   # processing.   return result   li = [elem_1,elem_2,elem_3]   process_list(li) 

now, there occasions might want process many 1d-lists. in these instances, there many 1d-lists in matrix-like structure (two-dimensional list or, if prefer, list of 1d-lists). each row represented 1d-list.

from design point of view, build on existing function process_list(li) so:

 def process_many_lists(many_li):   results = [ process_list(row) row in many_li ]   return results   many_li = [ \             [elem_11,elem_12,elem_13], \             ..., \             [elem_n1,elem_n2,elem_n3] \            ]   process_many_lists(many_li) 

alternatively, remove original process_list(li) function , refactor so:

 def process_any_amount_of_lists(li_or_many_li):   results = []   row in li_or_many_li:    # processing.   return results   li_or_many_li = [ \                   [elem_11,elem_12,elem_13], \                   ..., \                   [elem_n1,elem_n2,elem_n3] \                  ]   process_any_amount_of_lists(li_or_many_li) 

the first approach allows me keep separate dedicated function instances 1 list processed (these instances occur) through process_list(li), but, arguably, it's redundant. in addition, convolutes process_many_lists(many_li) function because works calling process_list(li).

the second approach using process_any_amount_of_lists(li_or_many_li) removes separate dedicated function instances 1 list processed, but, perhaps eradicates redundancy , convolution previous approach not rely on process_list(li). it's worth noting that, using approach, instances 1 list processed, 1 list must still presented function 2d list (e.g. li = [[elem_1,elem_2,elem_3]] not li = [elem_1,elem_2,elem_3]). of course, additional code include preliminary step perform different actions depending on dimensionality of list, situation having 2 separate functions anyway.

• is there clear answer here route should taken in scenario this?
• perhaps alternative approach not described above preferred?
• is solution exclusively "pythonic" or more general?